Multivariate Bayesian regression applied to the problem of network security

An Erratum has been published for this article in Journal of Forecasting 23(6): 461 (2004) . This paper examines the problem of intrusion in computer systems that causes major breaches or allows unauthorized information manipulation. A new intrusion‐detection system using Bayesian multivariate regression is proposed to predict such unauthorized invasions before they occur and to take further action. We develop and use a multivariate dynamic linear model based on a unique approach leaving the unknown observational variance matrix distribution unspecified. The result is simultaneous forecasting free of the Wishart limitations that is proved faster and more reliable. Our proposed system uses software agent technology. The distributed software agent environment places an agent in each of the computer system workstations. The agent environment creates a user profile for each user. Every user has his or her profile monitored by the agent system and according to our statistical model prediction is possible. Implementation aspects are discussed using real data and an assessment of the model is provided. Copyright © 2002 John Wiley & Sons, Ltd.     

From the history of low temperature physics: Prejudicial attitudes that hindered the initial development of superconductivity theory

In this paper we primarily study the difficulties in the application of the thermodynamical approach to the phenomenon of superconductivity, resulting from a specific interpretation of the low temperature behavior of the Maxwell equations. An attempt is made to elucidate the situation during the first period of the development of superconductivity theory and to render explicit prejudiced attitudes reminiscent of Feyerabend's natural interpretations.     

Understanding macroscopic quantum phenomena: The history of superfluidity 1941–1955

In this paper we attempt to investigate the historical and methodological aspects of the developments related to superfluid helium, concentrating on the period between 1941 and 1955. During this period, the various developments constituted a series of steps towards redefining and refining the two-fluid concept devised to explain the unexpected macroscopic behaviour of superfluid helium. The idea that superfluids are essentially ‘quantum structures on a macroscopic scale’ functioned as a heuristic principle which guided the theoretical physicists engaged in the above research programme.     

Genome-wide association scan identifies a colorectal cancer susceptibility locus on 11q23 and replicates risk loci at 8q24 and 18q21

In a genome-wide association study to identify loci associated with colorectal cancer (CRC) risk, we genotyped 555,510 SNPs in 1,012 early-onset Scottish CRC cases and 1,012 controls (phase 1). In phase 2, we genotyped the 15,008 highest-ranked SNPs in 2,057 Scottish cases and 2,111 controls. We then genotyped the five highest-ranked SNPs from the joint phase 1 and 2 analysis in 14,500 cases and 13,294 controls from seven populations, and identified a previously unreported association, rs3802842 on 11q23 (OR = 1.1; P = 5.8 x 10(-10)), showing population differences in risk. We also replicated and fine-mapped associations at 8q24 (rs7014346; OR = 1.19; P = 8.6 x 10(-26)) and 18q21 (rs4939827; OR = 1.2; P = 7.8 x 10(-28)). Risk was greater for rectal than for colon cancer for rs3802842 (P < 0.008) and rs4939827 (P < 0.009). Carrying all six possible risk alleles yielded OR = 2.6 (95% CI = 1.75-3.89) for CRC. These findings extend our understanding of the role of common genetic variation in CRC etiology.     

Covariance estimation for multivariate conditionally Gaussian dynamic linear models

In multivariate time series, estimation of the covariance matrix of observation innovations plays an important role in forecasting as it enables computation of standardized forecast error vectors as well as the computation of confidence bounds of forecasts. We develop an online, non‐iterative Bayesian algorithm for estimation and forecasting. It is empirically found that, for a range of simulated time series, the proposed covariance estimator has good performance converging to the true values of the unknown observation covariance matrix. Over a simulated time series, the new method approximates the correct estimates, produced by a non‐sequential Monte Carlo simulation procedure, which is used here as the gold standard. The special, but important, vector autoregressive (VAR) and time‐varying VAR models are illustrated by considering London metal exchange data consisting of spot prices of aluminium, copper, lead and zinc. Copyright © 2007 John Wiley & Sons, Ltd.     

Cooperative interactions between odorant-binding proteins of Anopheles gambiae

To understand olfactory discrimination in Anopheles gambiae, we made six purified recombinant OBPs and investigated their ligand-binding properties. All OBPs were expressed in bacteria with additional production of OBP47 in the yeast Kluveromyces lactis. Ligand-binding experiments, performed with a diverse set of organic compounds, revealed marked differences between the OBPs. Using the fluorescent probe N-phenyl-1-naphthylamine, we also measured the binding curves for binary mixtures of OBPs and obtained, in some cases, unexpected behaviour, which could only be explained by the OBPs forming heterodimers with binding characteristics different from those of the component proteins. This shows that OBPs in mosquitoes can form complexes with novel ligand specificities, thus amplifying the repertoire of OBPs and the number of semiochemicals that can be discriminated. Confirmation of the likely role of heterodimers was demonstrated by in situ hybridisation, suggesting that OBP1 and OBP4 are co-expressed in some antennal sensilla of A. gambiae.     

At the crossroads from Asia to Europe: spring migration of raptors and black storks in Dadia National Park (Greece)

Migrating raptors and black storks (Ciconia nigra) were studied in Dadia National Park, Greece, during spring 2003–2005. Vantage points and transects were used to evaluate magnitude, phenology, local variation and direction. We observed 23 species and 2030 individuals, including 715 common buzzards (Buteo buteo), 547 black storks and 136 short-toed eagles (Circaetus gallicus). Species-specific migration peaks were detected, starting in the second half of March, e.g. common buzzard, short-toed eagle, black stork, and ending early May, e.g. Levant sparrowhawk (Accipiter brevipes), red-footed falcon (Falco vespertinus), honey buzzard (Pernis apivorus). Most raptors were observed in the Evros valley, which may function as a migration corridor. The average direction of passing raptors was north. Species’ proportions and phenology in the study area were similar to those in neighbouring Bosphorus and Marmara Flyways. Further migration monitoring should be established in the area, which will provide important information not least to inform wind farm development location.